A Multiple Kernel Learning Approach for Air Quality Prediction

Abstract

Air quality prediction is an important research issue due to the increasing impact of air pollution on the urban environment. However, existing methods often fail to forecast high-polluting air conditions, which is precisely what should be highlighted. In this paper, a novel multiple kernel learning (MKL) model that embodies the characteristics of ensemble learning, kernel learning, and representative learning is proposed to forecast the near future air quality (AQ). e centered alignment approach is used for learning kernels, and a boosting approach is used to determine the proper number of kernels. To demonstrate the performance of the proposed MKL model, its performance is compared to that of classical autoregressive integrated moving average (ARIMA) model; widely used parametric models like random forest (RF) and support vector machine (SVM); popular neural network models like multiple layer perceptron (MLP); and long short-term memory neural network. Datasets acquired from a coastal city Hong Kong and an inland city Beijing are used to train and validate all the models. Experiments show that the MKL model outperforms the other models. Moreover, the MKL model has better forecast ability for high health risk category AQ.

Air Quality Prediction in Smart Cities Using Machine Learning Technologies Based on Sensor Data: A Review

“…A multiple kernel learning approach for air quality prediction [53]: Zheng et al proposed multiple kernel learning model with support vector classifier (MKSVC) as the base learner, which combines feature selection, metric learning and ensemble method for predicting air quality. For learning kernels, the centred alignment approach was applied, and for determining the optimal number of kernels, a boosting approach was applied.…”

A Self-Adaptive Multikernel Machine Based on Recursive Least-Squares Applied to Very Short-Term Wind Power Forecasting

“…MKL combines a set of kernels (basis kernels) in a linear, nonlinear or data-dependent way into a composite kernel, where the basis kernels can use different kernel functions or different values for the hyperparameters of a single kernel function [48]. Numerous studies have continuously improved the development of MKL applied in many subjects: classification of hyperspectral images [49], binary classification problems [50], air quality prediction [51], anomaly detection [52], object categorization [53], Alzheimer's disease diagnosis [54], oil painter recognition [55], multiclass classification [56], discriminating early-and late-stage cancers [57], subspace clustering [58], and many others.…”

“…The use of these success prediction functions as instancedependent weighing functions promotes locally discriminative base kernels while suppressing others. Zheng et al [51] introduced the multiple kernel support vector classifier, an MKL model, which embodies the characteristics of ensemble learning, kernel learning, and representative learning. The centered alignment approach is used to obtain the weight of each kernel, and a boosting approach is used to determine the proper number of kernels.…”

One Step Ahead Prediction of Ozone Concentration for Determination of Outdoor Air Quality Level

“…The proposed model shown high performance beyond the other models, also they praised SVM, RF, and ANN-MLP performance these presented very good performance than the sequence-to-sequence models LSTM and ARIMA. MKL-SVC model succeeded accuracy of 0.972 with MSE of 0.030 [17]. F. Martinez et al illustrated the potential to train a time-series dataset by KNN model.…”